An Optimal Pufferfish Privacy Mechanism for Temporally Correlated Trajectories

Temporally correlated trajectories are ubiquitous, and it has been a challenging problem to protect the temporal correlation from being used against users' privacy. In this paper, we propose an optimal Pufferfish privacy mechanism to achieve better data utility while providing guaranteed privacy of temporally correlated daily trajectories. First, a Laplace noise mechanism is realized through geometric sum of noisy Fourier coefficients of temporally correlated daily trajectories. Then, we prove that the proposed noisy Fourier coefficients' geometric sum satisfies Pufferfish privacy, i.e., the so-called FGS-Pufferfish privacy mechanism. Furthermore, we achieve better data utility for a given privacy budget by solving a constrained optimization problem of the noisy Fourier coefficients via the Lagrange multiplier method. What is more, a rigorous mathematical formula has been obtained for the Fourier coefficients' Laplace noise scale parameters. At last, we evaluate our FGS-Pufferfish privacy mechanism on both simulated and real-life data and find that our proposed mechanism achieves better data utility and privacy compared with the other state-of-the-art existing approach.